Taxonomic revision of the Andean genus Xenophyllum (Compositae, Senecioneae)

Abstract The Andean genus Xenophyllum (Compositae, Senecioneae) is distributed along the high-Andes from northeastern Colombia to northern Chile and northwestern Argentina, mainly thriving in the paramo and puna ecoregions. It comprises suffruticose plants forming dense mats, hummocks, or clumps of erect stems. They are characterized by displaying involucral bracts fused at the base, supplementary bracts absent, and mostly radiate capitula with white ray corollas, seldom yellow or pink (disciform in one species). Traditionally, Xenophyllum species were treated as members of the genus Werneria, a morphologically close genus that includes rosettiform or scapiform perennial herbs. As currently circumscribed, Xenophyllum mostly differs from Werneria in having elongate stems. Herein, the first modern and comprehensive revision of the genus recognizing twenty-two species and two subspecies is presented. Werneria decumbens is synonymized with X. weddellii, as well as X. fontii with X. humile and X. oscartovarii with X. dactylophyllum. Likewise, four varietal names and two sectional names are proposed as new synonyms. Seven names are lectotypified, the name X. sotarense is epitypified, W. decumbens neotypified, and the supraspecific name W. sect. Integrifoliae Rockh. is typified. The combination X. crassum subsp. orientalecomb. nov. is made. Descriptions and distribution maps are provided for all accepted species, in addition to an identification key. Ten species are illustrated, three of them for the first time.


Introduction
Xenophyllum V.A. Funk (Compositae, Senecioneae), is a genus of some 22 species distributed along the highlands of the Andes from northeastern Colombia to northwestern Argentina and northern Chile. The genus extends in distribution from the Colombian department of Norte de Santander (paramo del Almorzadero, ca. 7°N) to La Rioja Province in northwestern Argentina (Sierra de Famatina, ca. 29°S). Peru and Bolivia harbor the highest species diversity (12 and 11 species, respectively), followed by Ecuador (7 spp.), Chile (6 spp.), Argentina (5 spp.), and Colombia (3 spp.). The largest number of endemic species is found in Ecuador, 4 of the 7 species (57%) in the country (Fig. 1). Species of Xenophyllum are mostly small suffruticose plants characterized by displaying involucral bracts fused at the base, supplementary bracts absent, radiate capitula (disciform in one species), usually white ray corollas (yellow or pink in a few species), balusterform filament collars, obtuse or auriculate anther bases, truncate style branches with a crown of sweeping trichomes or penicillate, and glabrous or white-villous achenes. Two main habits can be differentiated according to the type of growth form (1) species forming dense mats or hummocks; (2) species forming clumps of somewhat distantly spaced stems or even having a shrubby habit (Calvo and Funk 2020).
Species of the genus were traditionally treated under the morphologically similar genus Werneria Kunth until the end of 20 th century, when Funk (1997a) coined the genus Xenophyllum. This new genus was meant to segregate from Werneria those species forming loose or tightly compressed hummocks or well-developed mats, and accordingly, the genus Werneria was re-circumscribed to accommodate the rosettiform species growing solitary or in small clumps. However, the type of growth form is not unequivocal for segregating both genera because some species of Werneria can also form compressed mats (e.g., W. aretioides Wedd., W. weberbaueriana Rockh.). We did not find any diagnostic morphological synapomorphies to support either of the two genera, but the presence or absence of elongate stems allows placing most species in one genus or the other. On this basis, Werneria comprises rosettiform or scapiform perennial herbs while Xenophyllum includes suffruticose species. Although phylogenetic studies are required to elucidate their evolutionary relationships, Funk (1997a) pointed out that these groups are probably not monophyletic. For the time being, and in pursuit of pragmatism, the revision of both groups has been presented separately following Funk's proposal. Indeed, her treatment was adopted in all subsequent regional catalogues and floras of the Andean countries (Nordenstam 1999;Jørgensen 2014;Freire and Ariza-Espinar 2014;Ávila et al. 2016). The taxonomic history of the group before the establishment of the genus Xenophyllum is detailed in Calvo et al. (in press). After the original publication of Xenophyllum in 1997, the taxonomic contributions into the understanding of the genus were limited to a few new combinations (Cabrera and Freire 1999;Calvo et al. 2017), a synopsis of the Peruvian species (Beltrán 2016), and the description of new species (Linares et al. 2010;Calvo and Funk 2020). For this reason, we present the first modern and comprehensive revision of the genus Xenophyllum. It includes updated nomenclature, identification keys, detailed descriptions, distributions maps, and drawings.
A comprehensive synonymy of the genus Xenophyllum was compiled. Types of all accepted names and synonyms were studied. However, we did not locate original material of Werneria dactylophylla var. glanduloso-denticulata Rockh. This appears in Unverified names and we suggest its application based on the opinion of previous authors.
A general description of the genus and detailed species descriptions were prepared. For this, qualitative characters were studied with the aid of a binocular dissecting microscope, while quantitative characters were recorded using a Mitutoyo digital caliper, CD-15DC. A Zeiss Standard 16WL microscope was used for examination of the achene trichomes. Information concerning the habitat, elevation, and flowering period of each species was obtained from the herbarium specimen labels.
Accepted species are presented in an order that represents their morphological affinities. An index to the taxonomic names treated is provided in Appendix I. Likewise, all studied exsiccatae can be found in Appendix II. The information of the examined specimens that do not correspond to type material is detailed at the end of each species. They are consecutively listed in alphabetical order by country, primary political division, and collector surname. When herbarium specimens from a political division have not been studied but the presence of the respective species is expected there, the name of that political division is included and marked accordingly in the Distribution and habitat section. The maps were generated using QGIS 3.4 Madeira. The country abbreviations used in the distribution maps are: ARG (Argentina), BOL (Bolivia), BR (Brazil), CH (Chile), COL (Colombia), ECU (Ecuador), and PER (Peru).

Morphology
Habit. Species of the genus Xenophyllum are small suffruticose plants with two main habits according to the type of growth form. One comprises the species forming dense mats or hummocks ( Fig. 2A, B), which exhibit the morphology and growth typical of cushion plants. They consist of a compact mass of closely spaced stems with reduced apical dominance. Underneath the living distal stem, the leaves wither and remain attached to the stem (marcescent leaves) or fall off and only the leaf base remnants persist (Fig. 2B). This nonphotosynthetic stem part resembles a rhizome, hence it is here referred to as rhizome-like stem and is described separately from the aerial photosynthetic part (Fig. 3A).
In contrast, the second habit type includes the species forming clumps of somewhat distantly spaced stems that are uniform in appearance along its whole length (Fig. 2C, D). Among these species, X. staffordiae (Sandwith) V.A.Funk can be considered shrub as the stems reach a height of ca. 60 cm.
Leaves. The leaves are alternate, simple, usually subimbricate to stellate-imbricate (spirally arranged), and are extended into a sheath-like base that is more or less noticeable depending on the amount of indumentum at the leaf attachment. The leaf length provided in the text strictly corresponds to the leaf lamina (sheath-like base excluded, Fig. 3C). The shape of the leaf lamina can be linear, triangular, or spatulate. A single Figure 2. Main habit types in Xenophyllum A, B species forming dense mats or hummocks C, D species forming clumps of somewhat distantly spaced stems A X. humile (Ecuador, Pichincha, pr. Papallacta) B X. sotarense (Colombia, Cauca, Sotará Volcano) C X. juniperinum (Chile, Antofagasta, Aucanquilcha Volcano) D X. digitatum (Bolivia, Potosí, cordillera Kari Kari). Pictures by Joel Calvo. species, X. esquilachense (Cuatrec.) V.A.Funk, has somewhat distantly arranged leaves that lack sheath-like base and are abruptly narrowed at the base.
Species of Xenophyllum are largely glabrous with an exception of a few species that can have floccose-lanate indumentum (e.g., X. staffordiae) or evanescent arachnoid trichomes (e.g., X. amblydactylum (S.F.Blake) V.A. Funk). The presence of indumentum is very variable at specific level, and therefore, it is not useful for distinguishing species.
Capitula. All species have radiate and erect capitula except for X. esquilachense that displays disciform capitula and X. staffordiae that has radiate capitula rather nodding. Among the radiate species, most taxa have ray corollas conspicuously surpassing the involucre (e.g., X. marcidum). Three species, however, are characterized by their ray corollas not surpassing the involucre, i.e., X. ciliolatum, X. juniperinum (Hieron.) J.Calvo, and X. staffordiae. When the ray corollas are well-developed, this character is very noticeable.
The ray florets are pistillate and fertile. The color of the ray corollas is usually white, sometimes slightly purplish on the abaxial surface near the apex. Three species have yellow or pale yellow ray corollas and X. roseum (Hieron.) V.A.Funk is the single species in the genus with pink ray corollas. The corolla limb has 2 to 7 veins (sometimes unconspicuous) and it is subentire to 3-toothed at the apex.
The disc florets are hermaphroditic. The corollas are tubular, 5-lobed at the apex, yellow, whitish, creamy, or purplish.
Involucres. The involucres are cupuliform or narrowly cupuliform. They lack supplementary bracts (calyculus) and the involucral bracts are typically fused at the base. The number of involucral bracts, although quite constant within each species, is not very useful for distinguishing purposes because of the significant overlap among several species. Rather useful is the involucral bract length; the provided measurements strictly refer to its free part (Fig. 3B). Likewise, it has to be clarified that the involucre length corresponds to the distance from the apex of the involucral bracts to the peduncle attachment point. The involucre was found to be a defining character of the genus. . Diagram of measurements used in descriptions A1 aerial photosynthetic part of the stem (referred to as aerial part in the descriptions of the species forming dense mats or hummocks) A2 rhizome-like stem of the species forming dense mats or hummocks B free part of the involucral bracts (notice the typically fused base) C1 leaf lamina C2 sheath-like leaf base. Details taken from Alice Tangerini's illustrations.
Further variability has been found in the morphology of the style branch apex. In all species examined it is truncate but differs on its ornamentation; some species bear a crown of sweeping trichomes (e.g., X. juniperinum, X. sotarense (Hieron.) V.A. Funk) whereas others display a tuft with scanty or many longer trichomes, i.e., penicillate (e.g., X. digitatum, X. rosenii (R.E.Fr.) V.A. Funk). We found that this character presents intraspecific variability (e.g., X. dactylophyllum, X. rigidum (Kunth) V.A. Funk), hence it has not been considered valuable for taxonomic purposes. Salomón et al. (2016) stated similar conclusions after studying the floral microcharacters in 72 South American Senecio species and several species from other Senecioneae genera. They noticed transitional states in trichome length, making difficult the difference between the typical crown of short trichomes and the apices with a tuft of elongate trichomes.
Achenes. The achenes are homomorphic, cylindrical, 6 to 9-ribbed, mostly glabrous except in the Ecuadorian endemic species X. funkianum J. Calvo, X. rigidum, and X. roseum that are white-villous. The achenes of X. acerosum, another species endemic to Ecuador, are described as having scattered arachnoid trichomes but such observation is based on the study of immature achenes, and therefore, mature achenes are required for confirming it. An accurate description of the achene size could not be provided for this latter species and for a few other species due to a lack of collections with mature achenes.
In X. funkianum and X. rigidum, the achene indumentum is composed of twin filiform trichomes, not myxogenic, with acute to subacute, asymmetrical, slightly forked or not apex; see Calvo and Funk (2020, fig. 2D).
The pappus of all species is capillary, composed of barbellate, whitish to partially rose-or purple-colored bristles.
Thus far, X. acerosum is only known from the type locality in the paramo of Yacumbi, which is located above Saraguro on the border between the Ecuadorian provinces of Azuay and Zamora-Chinchipe.
Phenology. Collected in flower in September and November.  Etymology. The adjective acerosus -a -um means needle-shaped, and it refers to the leaf shape of this species.
Notes. Xenophyllum acerosum is unique in the genus in its distant, rather solitary, erect stems arising from long rhizomes and by its linear-acicular, aristate leaves. The stems only bear leaves in the upper part.
The species is very common along the paramos of Colombia and Ecuador, but it becomes very scattered towards its southern limit. No collections have been studied from the region comprised between central Peru (Junín) and northern Bolivia (La Paz).
Phenology. Flowering nearly all year round. Etymology. The epithet humile means low-growing, close to the ground. Notes. Xenophyllum humile (Kunth) V.A.Funk is a morphologically polymorphic species with rhizome-like stems covered with lanate indumentum and marcescent leaves, obtuse to truncate leaf apex (usually callous-like tipped), (8-)10 to 13 involucral bracts, and (8-)10 to 13 ray florets with white corollas. The leaves are usually bent backwards due to a sort of articulation at the inflexion point between the leaf lamina and the sheath-like base. It is also characterized by its glabrous achenes, although we studied a single specimen from El Cajas (Azuay, Ecuador) displaying achenes with indumentum (Ulloa & Minga 1393, HA).
The species shows high variability in leaf size and density, which led some authors to describe infraspecific taxa or even to propose distinct species. Werneria fontii Cuatrec. was described from Tolima (Cordillera Central, Colombia) mainly on the basis of the compact habit of the plants thriving in this area. These plants are certainly smaller than typical forms and display short leaves with the leaf lamina barely spreading from the sheath-like base. Specimens also collected in the paramos of Tolima were previously used by Weddell (1856) to describe a variety that he characterized as having flesher leaves than typical variety, which is in line with the remarks stated by Cuatrecasas (1935) under W. fontii. However, these small and compact plants are not exclusive from this region. We studied similar forms from Farallones de Cali (Cordillera Occidental, Colombia), Nariño (southern Colombia), Tungurahua (Ecuador), and from Ancash (central Peru). On the other hand, some specimens from the paramos of Chingaza and Sumapaz (Cordillera Oriental, Colombia) display unusual narrow leaves that Cuatrecasas named in sched. as W. humilis var. angustifolia (Cuatrecasas 25628, US-01269081;Cleef 1316, US-01324876). From label information, it can be understood that these plants grow almost submerged in the water, which would explain the singular morphology of these forms. The Bolivian populations, which represent the southern limit of distribution of X. humile, rather exhibit a morphology similar to the typical forms. The puzzling pattern of the different morphologies along the whole distribution area makes unadvisable the recognition of more than a single species.
Xenophyllum humile shows morphological affinities with X. funkianum, X. rigidum, X. roseum, and X. sotarense (see comments under these species). Some specimens from Cundinamarca (Colombia) have the rhizome-like stems covered with a dense lanate indumentum and with scarce leaves along them (e.g., Pedraza & Franco 452,COL;Cleef & Florschütz 5469,COL,US). They might be confused with X. crassum subsp. orientale (Cuatrec.) J.Calvo, although their distributions areas do not overlap and X. humile has shorter and narrower leaves, shorter involucral bracts, and smaller capitula. We also studied a few specimens containing material mixed with Werneria pygmaea Gillies ex Hook. & Arn., a strictly rosettiform species that can be readily differentiated because it grows in lax clumps or solitary plants and its leaves are acute to obtuse at the apex and have the leaf lamina barely spreading from the sheath-like base (vs. leaves obtuse to truncate at the apex or callous-like tipped, usually with the leaf lamina spreading at nearly 90° from the sheath-like base in X. humile); see Cuatrecasas & Idrobo 27005 (COL) and Soderstrom 1293 (US).
An isolectotype of the name Werneria humilis Kunth is kept at the Willdenow Herbarium. This specimen was sent by Humboldt to Willdenow as stated at the bottom right of the sheet. It is not numbered but contains individuals identical to those from the lectotype at P.
Regarding the name Werneria fontii, it should be noted that the three specimens at MA are numbered with the same herbarium number, i.e., MA-246359. The isolectotypes, therefore, are indicated as MA-246359b and MA-246359c in order to differentiate them from the lectotype.
Phenology. Flowering from July to December. Etymology. The epithet sotarense refers to the Sotará Volcano, the locus classicus of this species. It is located in southern Colombia on the border between the departments of Cauca and Huila.
Notes. Xenophyllum sotarense is characterized by its linear, 4-5 mm long, densely stellate-imbricate leaves, which are typically aristate at the apex. The arista measures up to 0.5 mm long and falls as the leaf ages. The sheath-like base bears scattered multicellular trichomes.
It is morphologically close to X. humile, especially to those small and compact plants with leaf laminas barely spreading from the sheath-like bases (see comments under it). To avoid misidentifications the following characters should be carefully studied: leaf apex (aristate or clearly acute when the arista falls in X. sotarense vs. obtuse to truncate in X. humile) and stem indumentum (sparsely pilose in X. sotarense vs. lanate in X. humile). Xenophyllum sotarense also shows morphological affinities with X. roseum. The involucre length (4.9-5 mm in X. sotarense vs. 10-12 mm in X. roseum), the ray corolla length and color (6.3-6.5 mm, white in X. sotarense vs. 11.8-19 mm, pink in X. roseum), and the achene indumentum (glabrous in X. sotarense vs. white-villous in X. roseum) are useful characters for differentiating them.
The lectotype of Werneria sotarensis Hieron. at MA consists in a single sterile fragment barely informative and no isolectotypes appear to be extant; the duplicate at B was destroyed in 1943 (F0BN015821). For that reason, an epitype collected in the locus classicus has been designated. It is noteworthy that the plants from this locality are restricted on the volcano summit and co-occur with X. humile, a species much more abundant in the area that was also found at lower elevations.
The isotypes of Werneria leucobryoides S.F.Blake located at the Ecuadorian herbaria Q and QPLS show that the collector of this material was the Italian naturalist Enrico Festa instead of Luis Sodiro as the protologue indicates. This correction has been adopted since these specimens are the only ones that bear a label handwritten by Sodiro. It is feasible to think that the collector information was corrected on the specimens belonging to Sodiro's personal herbarium but omitted on the duplicates to be distributed.
Xenophyllum roseum is a narrow endemic species only known from El Cajas National Park (Azuay) and the Culebrillas Lagoon (Cañar). A single collection from Loja (Jameson s.n., US) would also support its presence in this province. However, the label information does not provide a precise locality ("province of Loxa"), and therefore, it has not been represented on the map. New collections are required to confirm that X. roseum thrives in Loja.
Phenology. Flowering from August to March. Etymology. The adjective roseus -a -um means rose-like, rose-colored, pink-colored. It refers to the characteristic pink ray corollas of this species.
Notes. Xenophyllum roseum is well defined by its linear, stellate-imbricate leaves that usually have the leaf lamina spreading at nearly 90° from the sheath-like base, and by the obtuse leaf apex that usually bears a quickly deciduous arista when young. The achenes are covered with a white-villous indumentum and the ray corollas are pink. This latter character makes this species unique within the genus. The color is very noticeable on living plants but in old dried specimens it might be less obvious, and therefore, somewhat equivocal for identification purposes.
It is morphologically similar to X. humile and X. sotarense. From the former species, it differs in the ray corolla length and color (11.8-19 mm, pink in X. roseum vs. 4.2-12.2 mm, white in X. humile), disc corolla length and color (5.7-8 mm, yellowish in X. roseum vs. 3.1-5 mm, whitish in X. humile), and achene indumentum (whitevillous in X. roseum vs. glabrous in X. humile). The deciduous arista that usually adorns the young leaves of X. roseum also helps to differentiate it from X. humile. With respect to X. sotarense, the differences are detailed under this latter species. Among the aforementioned species, X. roseum overlaps its distribution area with that of X. humile.
The species is known from a few collections near the Chimborazo Volcano. Its distribution area partially overlaps with those of X. humile and X. rigidum. See Calvo and Funk (2020) for further details.
Phenology. Flowering from April to July. Etymology. The epithet honors the American botanist Vicki A. Funk (1947Funk ( -2019, who described the genus Xenophyllum and greatly contributed to the understanding of the family Compositae.
Notes. Xenophyllum funkianum is distinguished by its creeping rhizome-like stems 20-35 cm long, the straight linear leaves extending into a sheath-like base that bears arachnoid trichomes, the dark-burgundy sessile involucres with 13 to 14 involucral bracts, the 12 to 13 ray florets with white corollas somewhat purplish beneath, and by having white-villous achenes.
Phenology. Flowering from April to November. Etymology. The epithet rigidum means stiff, inflexible and it describes the leaves of this species.  Notes. Xenophyllum rigidum is readily distinguished by its tough, linear, and apically obtuse leaves, as well as by its white-villous achenes. The rhizome-like stems are very robust and covered with a matted lanate indumentum where the leaf bases are sunked in. Another characteristic feature of this species is that the capitula are completely enclosed among the leaves.
Phenology. Flowering nearly all year round. Etymology. The adjective crassus -a -um means thick and it refers to the robust rhizome that this species displays.
Notes. This taxon forms dense mats or hummocks and it is characterized by its linear, apically obtuse leaves, which are flat to elliptical in cross section and prominently 1-nerved on both faces. The rhizome-like stems are robust, long, and covered with a matted lanate indumentum that covers the leaf bases.
Phenology. Flowering nearly all year round. Etymology. The adjective orientalis -is -e means eastern, which refers to the further eastern distribution of this taxon in relation to the typical subspecies. It occurs in the northern part of the Cordillera Oriental of Colombia (Eastern Ranges).
Notes. Xenophyllum crassum subsp. orientale is characterized by the leaf length (6.5-12.7 mm), the number of involucral bracts (11 to 13), and the number of ray florets (13 to 15). In dried specimens, the leaves are rather shiny. This taxon was synonymized with X. crassum subsp. crassum by Ávila et al. (2016) but no discussion supporting such taxonomic decision was provided. The aforementioned morphological differences against the typical subspecies, in addition to the fact that their distribution areas are well defined and do not overlap, lead us to maintain the treatment by Cuatrecasas (1980).
It should be noted that X. crassum subsp. orientale has been confused with Werneria pygmaea, a species that grows in the same habitat. Indeed, we detected several specimens that contain mixed material, including an isotype (see comments below). Xenophyllum crassum subsp. orientale has erect, robust rhizome-like stems whereas W. pygmaea is a strictly rosettiform species not forming hummocks. Further differences are detailed in Calvo et al. (in press). Some confusion with X. humile is also likely (see comments under it).
The presumed isotype of W. crassa subsp. orientalis Cuatrec. kept at COL (COL-000005514) comes from "extremo S en Peralonso, 3100 m", a locality slightly different from that indicated in the protologue, i.e., "vertiente norte, 3700−3800 m" [as "3600−3800 m" on the labels]. It is excluded from the type material as it does not belong to the same gathering. On the other hand, it has to be noted that the isotype at BC contains an individual corresponding to W. pygmaea (above on the right hand, with fragments of Disterigma sp. (Ericaceae) at the base).
Phenology. Flowering from February to September. Etymology. The adjective marcidus -a -um means withered. It probably refers to the fact that a large proportion of the stem bears marcescent leaves and the green ones are restricted to the distal part.
Notes. Xenophyllum marcidum is characterized by its triangular, densely imbricate leaves. Among the Xenophyllum members forming dense mats or hummocks, this species is readily distinguishable due to the leaves are denticulate in the upper half, acute at the apex, and rather flat in cross section. The capitula are radiate and have 14 to 21 ray florets with white, considerably long corollas.
Because of the marginally denticulate leaves, on dried specimens X. marcidum might be confused with X. ciliolatum, a species that has a partially overlapping distribution area. They can be easily differentiated by the habit (forming dense mats or hummocks in X. marcidum vs. forming clumps of erect or decumbent stems in X. ciliolatum), leaf lamina shape (triangular in X. marcidum vs. linear in X. ciliolatum), number of involucral bracts (13 to 15 in X. marcidum vs. ca. 8 in X. ciliolatum), number of ray florets (14 to 21 in X. marcidum vs. 8 to 11 in X. ciliolatum), and by the ray corolla length and color (7.4-12.9 mm, white in X. marcidum vs. 6.2-7.3 mm, yellow in X. ciliolatum).

Werneria ciliolata
Phenology. Flowering from February to November. Etymology. The adjective ciliolatus -a -um is the diminutive of ciliatus -a -um, which describes the denticulate leaf margin of this species.
Notes. Xenophyllum ciliolatum is one of the species displaying yellow ray corollas not surpassing the involucre. It is characterized by having erect or decumbent stems 4-7 cm tall, linear leaves 4.9-8.6 mm long with denticulate margin, ca. 8 involucral bracts, and 8 to 11 ray florets. The leaves are erect (not adpressed to the stem) and tend to be clustered upwards.
The morphologically closest species is X. juniperinum, known from southern Peru (Tacna), northern Chile, and western Bolivia (Oruro and Potosí). They differ in habit (erect or decumbent stems 4-7 cm long in X. ciliolatum vs. erect stems 8-15 cm long in X. juniperinum), leaf disposition (clustered up the stem and erect in X. ciliolatum vs. uniformly arranged along the stem and adpressed in X. juniperinum), leaf lamina shape (linear in X. ciliolatum vs. linear-triangular in X. juniperinum), leaf margin (denticulate in X. ciliolatum vs. minutely, irregularly denticulate in X. juniperinum), and involucral bract length (4.1-7.7 mm in X. ciliolatum vs. 2.8-3.4 in X. juniperinum). Their distribution areas do not overlap. Xenophyllum ciliolatum also shows morphological affinities with X. marcidum (see comments under it) and X. weddellii. From this latter species, X. ciliolatum differs in having denticulate leaf margin (vs. entire in X. weddellii), yellow ray corollas not surpassing the involucre (vs. white and conspicuously surpassing the involucre in X. weddellii), and disc corollas 5-7.3 mm long (vs. 4.5-5.2 mm long in X. weddellii). These species co-occur in southern Peru (Arequipa, Moquegua, Puno). The lectotype designation of the name Werneria ciliolata A.Gray by Calvo et al. (2017) should be rather attributed to Funk (1997a) although she used the term "holotype". It is here corrected to lectotype because it cannot be established that the author only used this element and that the gathering is represented by a single specimen (McNeill 2014).

Werneria juniperina
Most populations of this species are located between eastern Tacna (Peru), northeastern Arica-Parinacota (Chile), and northwestern Oruro (Bolivia). However, a few collections are from further south (northeastern Antofagasta in Chile and western Potosí in Bolivia). Its presence in the Chilean region of Tarapacá is expected.
Phenology. Flowering from April to December. Etymology. It is named after the genus Juniperus L. (Cupressaceae) due to the leaves of this species resembling those of some species belonging to the mentioned genus.
Notes. Xenophyllum juniperinum forms clumps of erect stems 8-15 cm long, which bear imbricate leaves that are adpressed to the stem and uniformly arranged along it. The leaves are somewhat broadened at the base and the margin is minutely, irregularly denticulate. The yellow ray corollas not surpassing the involucre is a distinctive character of this species, which is shared with X. ciliolatum (see comments under it for comparison purposes).
Phenology. Flowering nearly all year round. Etymology. This species honors H.A. Weddell (1819-1877), a British botanist (French by choice) and physician who devoted part of his life to the study of the South American flora.
Notes. Xenophyllum weddellii is a species forming clumps of rather decumbent stems with linear, entire, glabrous leaves. It has 8 to 12 involucral bracts and 9 to 15 ray florets with white corollas that conspicuously surpass the involucre. On living plants, the color of the leaves is typically glaucous green.
The leaf apex is acute but we studied a specimen showing two lateral divisions near the apex (Moreira- Muñoz & Diazgranados 2625, SGO). It is identified as X. weddellii because the remaining characters perfectly fit with this species. Two emerging protrusions near the apex were also observed in a few other collections, which indicates that Moreira-Muñoz's collection can be considered as an extreme of variability.
This species is morphologically similar to X. ciliolatum (see comments under it) and X. incisum. From this latter species, it mainly differs in the leaf lamina shape (linear-subulate, tapering upwards in X. weddellii vs. linear-oblong, similar width from base to apex in X. incisum) and leaf apex (acute in X. weddellii vs. 3-notched with the lobes rather truncate at the apex in X. incisum). It is interesting to mention that both species have similar ecological requirements and they usually thrive near the banks of the salt lagoons and moist valley bottoms. They could be referred to as allopatric species since their distribution areas are well defined and do not overlap.
The single original collection that we found comes from the Huasco Lagoon (Tarapacá, Chile). However, Philippi (1891) indicated additional sites in the protologue, i.e., "Machuca" and "regione inter Copacoya et Inacaliri". These localities are located further south in the region of Antofagasta, where this species appears to be absent. In contrast, X. incisum is not rare in this region and, indeed, the provenance of the type material is "Inter Copacoya et Inacaliri". The lack of original material of X. weddellii from these localities and the fact that we did not study any collection from there suggests that Philippi might had confused both species.
The holotype designation of the name Werneria weddellii Phil. has been corrected to lectotype according to McNeill (2014). In addition, the name Werneria decumbens Hieron. is synonymized with X. weddellii as suggested by Funk (1997a). Since the original material was likely destroyed at B in 1943, a neotype is designated. The provenance of the selected collection perfectly matches the protologue information. Philippi's collection at SGO as the first-step lectotype, designated as "holotype" by Funk (1997a: 239); SGO-060388! as the second-step lectotype, designated here; isolectotypes: LP-002607 (digital image!), SGO-000006431 (ruined!)).

Phenology. Flowering from October to March.
Etymology. The epithet incisum means sharply and deeply cut into, which describes the leaf apex of this species. However, its leaf apex is rather notched (shallowly cut into).
Xenophyllum incisum is somewhat variable in leaf apex shape. Typical forms have 3-notched leaf apex, usually with the central lobe wider than lateral ones, but in some specimens the apex is barely notched or even subentire (e.g., Moreira- Muñoz 298,CONC;Arroyo et al. 97315,CONC). In these specimens, however, the leaf apex is truncate and slightly curved forwards as typically in the species. Similarly, we studied one specimen having the leaf apex with each lobe 3-notched (Villagrán et al. 9321,CONC).
This species has morphological affinities with X. lorochaqui J.Calvo & V.A.Funk, X. weddellii (see comments under these species), and X. poposa. With regard to the latter species, the main and unequivocal character that allows a proper identification is the stem indumentum (glabrous in X. incisum vs. arachnoid in X. poposa). The characters concerning the leaf apex have traditionally been used for discriminating among these species, however, they are barely useful because X. poposa also shows high variability in leaf apex (see further comments under it). A rather useful character is the leaf section, which is obtusely triangular to slightly curved forwards in X. incisum vs. terete or almost so in X. poposa. The sheath-like base is also more broadened in X. incisum.
Phenology. Flowering nearly all year round. Etymology. The epithet poposa is the vernacular name of this plant as Philippi (1891) stated in the protologue "ubi Poposa vocabatur" [called poposa]. He also added "infusum plantae contra dolores colicos propinant" [for relieving stomach cramps and colic] referring to its medicinal properties. It is important to underline that poposa is a noun in apposition not to be declined. The correct epithet is, therefore, poposa instead of the widely misapplied poposum (Funk 1997a;Freire and Ariza-Espinar 2014;Beltrán 2016).
Notes. Xenophyllum poposa forms clumps of erect stems (rarely dense mats) characterized by having arachnoid indumentum. When the stems are very short and barely protrude from the soil, then, it tends to develop a mat-forming habit. The leaves are rather terete with the sheath-like base weakly broadened at the base. The leaf apex is usually undivided, rounded to truncate, and callous-tipped, however, plants having 2 to 3-notched apex are also found. It has 8 to 9 involucral bracts, 8 to 11 ray florets (with white corollas surpassing the involucre), 7 to 18 disc florets, and glabrous achenes. It is noteworthy that we observed a few ray florets with staminodes (Fig. 18G).
The type material of this species shows plants with clearly truncate and calloustipped leaf apex, which probably led Rockhausen (1939) to relate a collection displaying notched leaf apex to X. incisum and describe it as Werneria incisa var. pubescens Rockh. He differentiated the new variety from the typical one by the longer leaves and the pubescent-lanate sheath-like bases. A detailed study of the lectotype (CORD-00005636) reveals that the stem bears arachnoid indumentum and that the sheath-like bases are not conspicuously broadened. These characters are distinctive of X. poposa rather than X. incisum. Moreover, this latter species is restricted to the edges of salt lagoon edges and desert plains with a certain humidity. The type material of Rockhausen's variety was collected in the upper part of the Famatina Range (La Rioja, Argentina), which better matches the habitat of X. poposa (exposed slopes). On this basis, and because we believe that the glabrousness is a diagnostic character of X. incisum, we do not follow Rochkausen's treatment as previous botanists did (Cabrera 1948;Freire and Ariza-Espinar 2014). Otherwise, we consider it appropriate to recognize W. incisa var. pubescens as part of the variability encompassed by X. poposa. Our decision is also supported by the fact that we studied further specimens corresponding to X. poposa that display notched leaf apex. These populations appear very scattered from southern Peru to northwestern Argentina, namely in Moquegua, Oruro, Potosí, Tucumán, Catamarca, and La Rioja. In Moquegua some specimens are strikingly tiny, barely protruding from the soil, and exhibit 2 to 3-notched leaf apex (Beltrán 7736, USM;Funk et al. 13155, US, USM). Due to field work in this region we can state that these forms grow mixed with plants displaying undivided, rounded to truncate leaf apex. It also deserves to be underlined that the different leaf morphologies can even be found within the same individual. The specimen Funk et al. 13155 (US), for example, has undivided, 2-notched, and 3-notched leaves. In Potosí (Bolivia), we visited another population composed of midsize plants showing 3-notched leaves where each lobe can be, in turn, 2-notched (Calvo & Zárate 7869,BOLV,LPB;Cárdenas 362,US [see Fig. 19B]). No individuals with undivided leaves were found there. On the other hand, the specimens from Famatina (La Rioja, Argentina) are quite robust plants having stems 11-13 cm long and 2 to 3-notched leaves. Based on the study of herbarium specimens, we also can conclude that the populations having notched leaf apex from Oruro, Tucumán, and Catamarca grow mixed with or close to the typical plants. All these examples demonstrate that a great intrapopulation and interpopulation variability exists in terms of both leaf apex and habit. Our efforts for trying to propose an infraspecific classification in pursuit of recognizing the distinct morphologies failed because of the intermediate forms and the lack of defined distribution patterns. Therefore, we believe that X. poposa should be treated as a highly variable species.  The distribution area of this species partially overlaps with that of X. incisum in the region comprised between Antofagasta (Chile), Potosí (Bolivia), Salta, and Jujuy (Argentina). However, they occur in different ecological niches as already mentioned above. The morphological differences between them are commented under X. incisum. Likewise, X. poposa might be confused with X. lorochaqui (see notes under it).
Regarding nomenclatural issues, it is noteworthy that Funk's typification of the name Werneria poposa Phil. has been narrowed to a single specimen because two duplicates of the same gathering were found at SGO (Turland et al. 2018, ICN Art. 9.17). The specimen SGO-000006433 has therefore been selected as the second-step lectotype because it is most complete and best preserved.
The original material of Werneria lorentziana Hieron. comprises the syntypes Stübel 107 and Stübel 117, both apparently destroyed. Rockhausen (1939) designated as "typus" the collection Stübel 117 but indicated the locality that in the protologue corresponds to the number Stübel 107. Since these specimens are not extant anymore and we did not find pictures of them at F, such mismatch cannot be elucidated. Later, Freire and Ariza-Espinar (2014)  eronymus 117 (CORD). Although the collector number coincides with one of the syntypes, this element is not part of the original material and has to be treated as a neotype.
Phenology. Flowering from January to March. Etymology. The epithet lorochaqui is the vernacular name of this plant and it means parrot's foot ("loro" from Spanish: parrot; "chaqui" from Quichua: foot). It probably responds to the resemblance of the leaves with a parrot's foot.
Notes. Xenophyllum lorochaqui can be identified by the glabrous erect stems 10-20 cm tall (usually only bearing leaves in the upper part), leaf lamina 7.9-11.5 mm long, leaf apex 3-notched with the central lobe entire or barely notched and longer than the lateral ones, involucre with ca. 13 involucral bracts, and by displaying 26 to 39 ray florets with white corollas.
This species shows morphological affinities with X. incisum, X. dactylophyllum, and X. poposa. The differences against X. incisum are the leaf lamina length (3.2-7.3 mm vs. 7.9-11.5 mm in X. lorochaqui), the length of the leaf apex lobes (similar among them vs. central lobe longer than lateral ones in X. lorochaqui), the involucre length (6.4-8.3 mm vs 10.3-10.9 mm in X. lorochaqui), the involucral bract number (8 to 9 vs. ca. 13 in X. lorochaqui), and the ray floret number (8 to 13 vs. 26 to 39 in X. lorochaqui). With regard to X. dactylophyllum, the leaf apex is at least 9-divided (fingerlike) with the primary division extending deeper than the subsequent ones, whereas in X. lorochaqui the leaf apex is 3-notched with the central lobe longer than lateral ones. From X. poposa, the distinguishable characters are the stem indumentum (arachnoid in X. poposa vs. glabrous in X. lorochaqui), the leaf lamina length (2.5-7.1 mm in X. poposa vs. 7.9-11.5 mm in X. lorochaqui), the involucral bract number (8 to 9 in X. poposa vs. ca. 13 in X. lorochaqui), and the ray floret number (8 to 11 in X. poposa vs. 26 to 39 in X. lorochaqui). Among these species, the distribution area of X. lorochaqui partially overlaps with that of X. incisum and X. poposa. gensen (2014) also indicated this species from Oruro (Bolivia) on the basis of a misidentified collection that actually corresponds to X. digitatum (Liberman 352,LPB).
Phenology. Collected in flower in March and November. Etymology. It is named after Eric von Rosen (1879Rosen ( -1948, a Swedish aristocrat, explorer, and ethnographer. Notes. Xenophyllum rosenii is unique in the genus by its 2-forked leaves. The species has been confused with X. digitatum (e.g., Meneses 5446,LPB;Treviño 761,HSP;Tupayachi 736,US). A detailed study of the leaves highlights that these species can be straightforwardly discriminated from one another by the leaf apex (2-forked in X. rosenii vs. 3-forked in X. digitatum). The other characters appear to be useless because of a significant overlapping among them.
Since two specimens of the gathering Fries 862 were found at UPS, the lectotypification has been further narrowed to a single specimen by way of a second-step lectotype (UPS-V-833144).
Phenology. Flowering nearly all year round. Etymology. The adjective digitatus -a -um refers to the shape of the leaves, which have divisions arranged like those of a bird's foot (3-forked).
Notes. Xenophyllum digitatum is characterized by its 3-forked leaves, where each lobe can be, in turn, 2 to 3-divided. The depth of the primary division reaches 1/3-1/2 of the total leaf length and the leaf margin usually have some cilia in the lower half. It has quite large involucre (12.1-16.7 × 9.8-15.7 mm) with 12 to 13 involucral bracts, 21 to 36 ray florets (with white corollas conspicuously surpassing the involucre), 49 to 63 disc florets, and glabrous achenes.
Distribution and habitat. Endemic to Argentina (Jujuy, Salta). It grows in rocky slopes on rather wet soils of the dry puna ecoregion, between elevations of 3700-5125 m (Fig. 23).
Phenology. Flowering from January to May. Etymology. The epithet pseudodigitatum refers to the resemblance between this species and X. digitatum. It probably responds to the fact that both species have 3-forked leaves.
Notes. Xenophyllum pseudodigitatum is characterized by its 3-forked leaves. The leaf lobes are 1.3-2.6 mm long, which represents 1/8-1/7 of the total leaf length, and they are acute to mucronate at the apex (rarely notched); on living plants, they are noticeably spread out. The involucre length ranges from 7 to 8.4 mm and has 10 to 11 involucral bracts. The radiate capitula are composed of 20 to 25 ray florets and 29 to 33 disc florets.  This is a narrow endemic species to northwestern Argentina that has erroneously been reported from northern Chile (Ricardi and Marticorena 1966;Freire and Ariza-Espinar 2014;Moreira-Muñoz et al. 2016) and southern Peru (Gonzáles et al. 2016). All these records are confusions with X. digitatum. In order to avoid further misidentifications, the following characters should be studied: leaf lobe length (1.3-2.6 mm in X. pseudodigitatum vs. 3.6-5.9 mm in X. digitatum), involucre length (7-8.4 mm in X. pseudodigitatum vs. 12.1-16.7 mm in X. digitatum), number of involucral bracts (10 to 11 in X. pseudodigitatum vs. 12 to 13 in X. digitatum), ray corolla length (7.1-8.2 mm in X. pseudodigitatum vs. 9.2-14.1 mm in X. digitatum), and number of disc florets (29 to 33 in X. pseudodigitatum vs. 49 to 63 in X. digitatum). Their distribution areas do not overlap.
Phenology. Flowering from February to September. Etymology. The adjective decorus -a -um means handsome, elegant, decorous. Indeed, Blake (1928) pointed out in the protologue that it is "an attractive and very distinct species".
Notes. This is a distinctive species due to the shortly ciliate leaf margin and the 3-notched leaf apex (the lateral lobes being conspicuously larger than the central one). The leaf lamina is glabrous, linear-oblong, slightly broadened towards the apex, flat in cross section but tending to be curved forwards. It has 13 to 14 involucral bracts, 20 to 21 ray florets with white corollas, 37 to 58 disc florets, and glabrous achenes. Xenophyllum decorum has the leaf apex somewhat similar to that of X. incisum, however, their distribution areas are geographically very distant. In addition, these species differ in the leaf length (7.2-16 mm in X. decorum vs. 3.2-7.3 in X. incisum), leaf margin (shortly ciliate in X. decorum vs. not ciliate in X. incisum), number of involucral bracts (13 to 14 in X. decorum vs. 8 to 9 in X. incisum), number of ray florets (20 to 21 in X. decorum vs. 8 to 13 in X. incisum), and length of ray corollas (10.4-11.9 mm in X. decorum vs. 6.2-8.9 mm in X. incisum).
It has to be noted that we had problems in identifying Beltrán 7987 (USM). This specimen exhibits the characteristic flat and ciliate leaves of X. decorum but has the leaf apex strongly divided (ca. 2.5 mm deep) with each lobe entire or divided once again. It might be treated as an intermediate form between X. decorum and X. dactylophyllum or X. digitatum since these species co-occur in the region where Beltrán's specimen was collected.
Notes. Xenophyllum amblydactylum forms clumps of erect stems 4-7 cm tall that bear linear, 4.3-10 mm long leaves. The leaf apex is 3-notched and the lobes are small (0.8-1.4 × 0.6-1 mm) and connivent. The leaf lamina has arachnoid trichomes on the adaxial surface that are quicky evanescent. The involucre is 8.8-9.8 mm long and it is composed of 10 to 13 involucral bracts. The capitula are radiate and display ca. 11 ray florets with white corollas. This is a poorly known and collected species. Aside from the type material, we only studied one collection from southeastern Lima Department (Beltrán 4174, USM) and a specimen that was bought as medicinal in a farmers market in the city of Lima (Yarupaitán 1499b, USM). Further collections are required in order to improve the understanding of its variability and properly describe the achenes.
Xenophyllum datylophyllum has been reported from the Bolivian department of Potosí on the basis of the collection Steinbach s.n. (Jørgensen 2014;Tropicos 2020). This material, which has not been located, was collected nearby Andacaba, a locality ca. 30 km south of the city of Potosí. If the identification is confirmed, it would represent the southern distribution limit of this species. The specimen US-00974139 most likely corresponds to a mislabeling because both the locality (San Carlos de Bariloche, Río Negro Province, Argentina) and the elevation (600 m) do not correspond to the known distribution area and altitudinal rank of this species.
Phenology. Flowering nearly all year round. Etymology. The epithet dactylophyllum means finger-like leaves and it refers to the resemblance between the leaves of this plant and a hand. Notes. Xenophyllum dactylophyllum is characterized by having glabrous or floccoselanate finger-like leaves, with at least 9 divisions at the apex. The divisions are usually arranged in multiples of three, the primary one extending deeper than the subsequent ones. The capitula are radiate, erect, with 17 to 22 ray florets with white corollas conspicuously surpassing the involucre.
Another species with finger-like leaves is X. staffordiae, which has occasionally been confused with X. dactylophyllum. They can be readily differentiated by the habit (11-24 cm tall in X. dactylophyllum vs. 30-60 cm tall in X. staffordiae), leaf divisions (at least 9-divided, with the primary division extending deeper than the subsequent ones in X. dactylophyllum vs. at least 12-divided, with all divisions similar in length in X. staffordiae), capitulum position (erect in X. dactylophyllum vs. rather nodding in X. staffordiae), and length and color of the ray corollas (9.2-11 mm, surpassing the involucre, white in X. dactylophyllum vs. 6.4-8.3 mm, not surpassing the involucre, pale yellow in X. staffordiae). Their distribution areas overlap from central to southern Peru. Other morphologically similar species are X. amblydactylum and X. lorochaqui (see comments under these species).
Our efforts in locating the type material of X. oscartovarii E.Linares, J.Campos, Nauray, Vicente Orell. & A.Galán were unfruitful. Due to the original description and the pictures included in the publication (Linares et al. 2010), we believe that it fairly corresponds to X. dactylophyllum. Oddly, this latter species was not mentioned in the taxonomic discussion of the presumed new species. In addition, we studied several specimens of X. dactylophyllum collected close to the type locality of X. oscartovarii. Additional